1. Field of the Invention
The present invention relates to an image forming apparatus which includes a fixing portion for heating an unfixed toner by heat (heat-fixing), formed on a recording material thereon.
2. Description of the Related Art
As heating devices for a recording material, there are conventionally known various methods and configurations such as a heat-roller method, a hot-plate method, a heat-chamber method, and a film-heating method. Those heating devices all include heating elements (heat members). In order to maintain a temperature of the device at a predetermined temperature (predetermined image fixing temperature), the temperature is controlled by controlling power supply to the heating element.
Among such various conventional heating devices, the heating device of the film-heating type is highly effective and practical.
The heating device of the film-heating type includes a thin heat-resistive film, a driving means (unit) for the film, a heating element fixed and supported in the film, and a pressure member which is disposed oppositely to the heating element and bonds an image bearing surface of a recording material to the heating element through the film. At least during image heating, the film is moved in a forward direction at substantially the same speed as the recording material, which is conveyed in between the film and the pressure member, and the film passes through a nip portion formed as an image heating portion by a pressing portion between the heating element and the pressure member sandwiching the traveling film. A visible image bearing surface of the recording material is accordingly heated by the heating element through the film to fix a visible image by heat. Then, the film passing through the image heating portion is separated from the recording material at a separation point. This is a basic configuration of the heating devices. Such heating devices of a film-heating type can use a heating element having a low heat capacity and a high temperature-increase rate, and a thin film. Thus, power can be saved, and a shortened wait time (quick start) can be achieved. This type of the heating device is advantageous in eliminating various disadvantages of the other conventional heating devices, which is effective.
In recent years, heating devices have been proposed, which reduce uneven toner melting caused by roughness of a recording material by disposing an elastic layer in a heat film.
In temperature control in the heating devices of the film-heating type, in many cases, the output of a thermistor disposed on the heating element is subjected to A/D conversion, and captured by a CPU. Then, based on a comparison result of a detected temperature with a target temperature, referring to a predefined control table, power supply to the heating element is controlled by PID control for performing proportion (P) control, integral (I) control, and differential (D) control.
In this case, the control of power supply to the heating element is performed by turning an AC voltage ON/OFF through a triac. Wave number control or phase control is used for the power supply control. Power is minutely controlled by controlling a power supply ratio, thereby reducing the amplitude of the temperature of the heating element as much as possible.
The wave number control is ON/OFF control for each half wave, in which several waves of an input AC voltage are set as a predetermined cycle (one control cycle), and which wave is turned ON and which wave is turned OFF are determined for each predetermined cycle. In other words, the wave number control is a method of controlling a power supply ratio based on an ON/OFF duty ratio within the predetermined cycle.
For example, one half wave=10 milliseconds is set when a frequency of alternating current power is 50 Hz. When a predetermined cycle is 20 half waves=200 milliseconds=1 cycle, power supplied to the heating element is revised for every 20 half waves. The minimum power is full OFF (20 half waves full OFF), and the maximum power is full ON (20 half waves full ON). The amount of power supplied for each cycle is divided into 21 levels where 0 half wave to 20 half waves are ON.
In this control, when a waveform of the input AC voltage is as illustrated in FIG. 10, as an example, a current supplied to the heating element has a waveform illustrated in FIG. 11.
The phase-control method is a method of controlling a power supply angle within one wave of the input AC voltage. A current supplied to the heating element has a waveform illustrated in FIG. 12.
In the wave-number control method, the harmonic current is small, while flicker noise is large. In the phase-control method flicker noise is small, while the harmonic current is large.
In wave number control, the power supply ratio is controlled for each predetermined cycle of several waves, and hence a revising cycle must be prolonged to increase the contained number of waves in order to minutely control the power supply ratio. However, the power supply ratio is permitted to be revised for each predetermined cycle. Thus, when the revising cycle is excessively long, switching of the power supply ratio is delayed, disabling supply of appropriate power when necessary. Hence, the power supply ratio and the revising cycle must be set to be balanced with each other.
In phase control, one control cycle is one half wave and hence the power supply ratio is minutely controlled within one half wave, and the power supply ratio is revised for each one full wave at the minimum. Thus, in phase control, the power supply ratio, and more specifically, the power, can be revised more minutely, and temperature ripples of the heating element accompanying the control can be reduced. However, costs of the apparatus are higher in the case of phase control because a noise filter is necessary and a circuit configuration is complex. On the other hand, wave number control has no such cost increase.
Thus, the control is chosen according to apparatus requirements. In particular, in a recent case of using a commercial power source of 200 V, not the phase control but the wave number control is often employed in order to reduce the harmonic current.
Under those circumstances, for example, as disclosed in Japanese Patent Application Laid-Open No. H10-333490, there has been proposed an apparatus configured to switch wave number control and phase control between 200 V and 100 V according to an input AC voltage.
A method has been proposed, which combines phase control and wave number control, in which the phase control is used for at least one half wave within a revising cycle of the wave number control so that a harmonic current is reduced more than when only the phase control is used, and the revising cycle of a power supply ratio is set shorter than when only the wave number control is used to perform more minute control. As an example, refer to Japanese Patent Application Laid-Open No. 2003-123941.
In the heating devices of the film-heating type, especially in a device which includes an elastic layer in the heat film, entry of the recording material into a heat nip portion may be accompanied by an unstable heating state of the recording material. The unstable state occurs because, if the recording material enters a stable temperature state, heat is suddenly absorbed by the recording material immediately after the entry of the recording material into the heat nip, causing a sharp reduction in heat film temperature, and overshoot occurs subsequently when the temperature increases, resulting in great temperature fluctuation of the heat nip.
With regard to the improvement of this phenomenon, the inventors of the present invention have disclosed a method of correcting power supplied to the heating element before temperature fluctuation occurs due to the entry of the recording material in Japanese Patent Application Laid-Open No. 2004-078181.
After the entry of the recording material into the heat nip has been accompanied by the sharp reduction in temperature of the heat film, the temperature is kept low when this portion comes into contact with the recording material again after one rotation of the heat film. More specifically, a phenomenon occurs where the temperature of the heat film drops in a portion corresponding to second rotation of the heat film on the recording material, and image glossiness declines. Meanwhile, it is only an instant immediately after the entry of the recording material causing a sudden change of the heat state that the entry of the recording material causes a large reduction in temperature of the heat film. By the PID control, the heat state is soon stabilized to a certain level, and the temperature reduction is eliminated. Thus, it is only at a portion corresponding to a leading edge of the second rotation that image glossiness declines in the portion corresponding to the second rotation of the heat film on the recording material.
There is a great difference in image glossiness between the portion corresponding to the leading edge of the second rotation of the heat film and a portion corresponding to a trailing edge of the first rotation thereof, and hence a glossiness difference may clearly appear as a step on the boundary. This phenomenon is conspicuous especially when glossy paper is passed through the nip.
In order to reduce the glossiness difference, the power correction must be minutely controlled so that glossiness can be equal at joint portions of the first rotation and the second rotation. More specifically, the temperature reduction of the heat film in the portion corresponding to the leading edge of the second rotation must be complemented so that temperatures can be equal at the leading edge of the second rotation and the trailing edge of the first rotation even if heat is removed at the leading edge of the first rotation.
A mechanism of complementing the temperature reduction based on the power correction is as follows. First, the entry of the recording material causes a reduction in temperature of a heat film surface. Unless power correction is performed, as described above, the temperature of this portion is kept low, and a glossiness difference occurs after one rotation of the heat film. When power correction is performed to forcibly input predetermined power before the entry of the recording material, even if the temperature of the heat film surface drops once, the power forcibly input during one rotation, specifically, heat energy, is conducted to the heat film surface. The temperature reduction is canceled, and a predetermined temperature is restored when the leading edge of the second rotation of the heat film corresponding to the recording material entering portion of the heat film comes into contact with the recording material again.
As is obvious from this mechanism, a portion where the heat generated by the power correction warms an inner surface of the heat film must substantially completely match the portion where the entry of the recording material has caused the reduction in temperature.
Such a case requires accuracy stricter than when the temperature control is simply stabilized. In particular, in the case of a recording material such as glossy paper, sensitivity of glossiness to temperature is very high, and only a slight temperature difference appears as a glossiness difference, more specifically, a step of glossiness in this case. Hence, the control width of a desirable surface temperature is very small.
In order to set the temperature of the trailing edge of the first rotation to be equal to the leading edge of the second rotation, power correction for accurately compensating for the temperature reduction at the leading edge of the second rotation must be performed. High accuracy is required not only for setting the power, but also for timing of the power correction. This is because the step in the temperature occurs in accordance with a delta function and, in order to complement the temperature reduction so as to prevent the occurrence of the step, the power must be changed in a complementary manner at an accurate timing of a delta function with respect to timing of the occurrence of the step.
When power correction timing shifts even slightly from the appropriate correction timing, the temperature reduction cannot be adequately compensated for due to a shortage of power or due to excessively input power, causing a problem of hot offset. In other words, when timing of starting power correction shifts even slightly, effects of the power correction are reduced.
However, in the apparatus which employs the wave number control, correction cannot be performed at a timing to perform power correction in response to the entry of the recording material, and temperature fluctuation caused by the entry of the recording material cannot be sufficiently reduced.
The above-mentioned problems occur for the following reason. The revising cycle of the power supply ratio of the wave number control comprises several waves, and hence the revised frequency is small. Thus, there is almost no case where the revised timing matches the power correction timing.
FIG. 13 is a timing chart illustrating revising cycles of power supply ratios of wave number control and phase control and timing of recording material entry and power correction.
In this example, the revising cycle of a power supply ratio of the wave number control is 20 half waves. The timing charts show revised timing A of a power supply ratio of the wave number control, and revised timing B of a power supply ratio of the phase control. Power correction is performed at timing C, and a recording material enters the heat nip at timing D. In the example of FIG. 13, power correction is started 130 milliseconds before the entry of the recording material into the heat nip, and the power control is finished 30 milliseconds after the entry of the recording material into the heat nip.
In the wave number control, the revising cycle of the power supply ratio is long, and hence the shift of the timing for actual correction from the appropriate correction timing is large. In the illustrated example, the power supply ratio is controlled by 20 half waves, and hence there is a shift (delay) of a maximum of 20 milliseconds (in the case of 50 Hz) from issuance of a power correction start command to the actual execution of correction. In this case, the power correction period is 160 milliseconds, which is the sum of 130 milliseconds before the entry of the recording material and 30 milliseconds after the entry. Thus, when there is a maximum shift, power correction is started after the time of a power correction stop. More specifically, in actuality, a command of a power correction stop is issued before a power correction start, and hence no power correction is performed.
In the above-mentioned example, the power supply ratio is changed after the command of the power correction start is issued, and hence a shift of timing is in a direction where execution of correction is always delayed. On the other hand, the start timing of the power correction is known beforehand, and hence the maximum amount of shift can be somewhat reduced by performing correction when the revised timing of the power supply ratio comes at the closest timing before/after the start timing of the power correction based on the assumption of a shift. Even in this case, however, the amount of shift is ±100 milliseconds at a maximum with respect to appropriate power correction timing.
FIGS. 14 to 16 illustrate temperature states of the heat film surface when such a timing shift occurs. In a graph of each of FIGS. 14 to 16, a horizontal axis indicates time, and a vertical axis indicates a surface temperature of the heat film. FIG. 14 illustrates a case where power correction is performed at the appropriate timing. FIG. 15 illustrates a case where a power correction start shifts before the appropriate timing. FIG. 16 illustrates a case where a power correction start shifts after appropriate timing. The entry of the recording material into the heat nip causes a reduction in temperature of the heat film. In FIG. 14, the difference in surface temperature of the heat film before and after the entry of the recording material into the heat nip is suppressed to about Δ2 deg. In FIG. 15, the difference in surface temperature of the heat film before and after the entry of the recording material into the heat nip is Δ8 deg because the surface temperature greatly increases before the entry of the recording material into the heat nip. In FIG. 16, the difference in surface temperature is about Δ8 deg because the entry of the recording material into the heat nip causes a great reduction in surface temperature.
As is obvious from FIG. 15, when power correction is performed at a shifted timing, if correction is performed before the appropriate timing, the temperature of the heat nip increases too greatly, causing excessive heating. When the recording material bearing a toner image enters the nip, toner is melted excessively to generate a hot offset phenomenon. High power is supplied before the appropriate timing, and hence the temperature of the heat film becomes too high until the entry of the recording material, and the glossiness of the recording material becomes higher in a portion corresponding to a trailing edge of the first rotation of the film. Thus, a horizontal strip of uneven brightness occurs so as to emphasize a step (a difference in glossiness) between the trailing edge of the first rotation and the leading edge of the second rotation. On the other hand, if correction is performed after the appropriate timing as illustrated in FIG. 16, the reduction in the amount of heat caused by the entry of the recording material into the nip cannot be compensated for, greatly reducing the temperature. In this case, the glossiness of a portion corresponding to the second rotation of the heat film becomes too low, resulting in an uneven brightness where a step between the trailing end of the first rotation and the leading edge of the first rotation is clearly observed.
In order to deal with the problem, the revising cycle of the power supply ratio may be shortened. In such a case, the number of waves within the revising cycle is reduced, disabling minute setting of the power supply ratio, and temperature control is hindered.
Needless to say, a timing shift occurs also in the case of the phase control. A value of the shift is 1 full wave=20 milliseconds (in the case of 50 Hz) at a maximum. Even with a shift of this level, the influence is not necessarily nil. The inventors of the present invention have conducted a study, and found that uneven brightness is somehow within a permissible range with this amount of shift. In other words, unless the phase control is used, a level which permits a timing shift cannot be set.
However, the phase control has a problem of generating a harmonic current, and hence the phase control cannot always be employed as described above. Moreover, since Europe belongs to a 200 V zone and has strict rules on harmonic currents, the phase control cannot be used, and therefore wave number control must be used.
In the control in which the phase control is used for at least one half wave within a revising cycle of the wave number control of a power supply ratio, the revising cycle of the power supply ratio can be shortened, and thus there are some improvement effects for addressing the problem. However, if the number of waves within the revising cycle is reduced in order to shorten the revising cycle of the power supply ratio, the number of waves for performing the phase control relatively increases, increasing harmonic currents. If this phenomenon is prevented, the power supply ratio cannot be set minutely. A permissible level is reached only by using the phase control for all of the cycles as described above, and hence there is a limit on improvement.